Heavy drinking is widespread and causes serious medical and social problems, including alcohol dependence. Treatment for non-dependent heavy drinkers now relies on simple advice (SA) and brief counseling (BC). The oral opiate antagonist naltrexone (NTX) has been effective in reducing heavy drinking by alcohol-dependent subjects in two studies, particularly when coupled with coping skills training. NTX reduced the desire to drink. We want to know if NTX and brief counseling yielded a significant reduction in total alcohol consumption, drinking frequency, heavy drinking, desire to drink, significant reduction in total alcohol consumption, drinking frequency, heavy drinking, desire to drink, and serum GGT levels in heavy drinkers. Many subjects' preference for sweet/rich foods declined during treatment, suggesting that opiate antagonism may more generally reduce appetitive drives. Alcohol and food consumption are interrelated and are modulated by brain opiate systems. We now propose a six-week, placebo-controlled, double-blind, randomized trial of NTX (50 mg/day) in 316 non-dependent heavy drinkers recruited from general medical clinics and via media advertisement. Males must consume 28 drinks/week, and females 18 drinks/week, or subjects must meet DSM-III-R SA. The others will receive four BC sessions providing drink reduction skills training. A 1-year post-treatment period will evaluate the stability of drinking reduction. This large sample size will provide adequate power to detect main and interactive effects of medication (NTX/PBO) and counseling (SA/BC) condition on drinking outcomes. Primary outcomes will be alcohol consumption and the frequency of heavy drinking days. Secondary outcomes will include adverse effects; subjects; confidence in resisting heavy drinking, desire to drink and consume sweet and rich foods; and consumption and preferences for sweets and rich/fatty foods. We hypothesize that: 1) NTX will be superior to PBO in reduction of desire for, and consumption of alcohol; 2) NTX and PBO will be equally well-t operated; 3) the efficacy of NTX will be enhanced by BC, which will be more effective than SA in facilitating drinking reduction; 4) NTX will be superior to PBO in reducing the desire for, and consumption of sweet/rich foods; and 5) these effects will persist during the 12-month post-treatment period. This study will test the efficacy of two low-cost promising treatments for non-dependent heavy drinkers. It will also evaluate the mechanisms of action of NTX and counseling methods. Its findings will be relevant to treatment of heavy drinker in general medical settings.